Maternal aryl hydrocarbon receptor activation protects newborns against necrotizing enterocolitis

Abstract

Necrotizing enterocolitis (NEC) is a disease of premature infants characterized by acute intestinal necrosis. Current dogma suggests that NEC develops in response to post-natal dietary and bacterial factors, and so a potential role for in utero factors in NEC remains unexplored. We now show that during pregnancy, administration of a diet rich in the aryl hydrocarbon receptor (AHR) ligand indole-3-carbinole (I3C), or of breast milk, activates AHR and prevents NEC in newborn mice by reducing Toll-like receptor 4 (TLR4) signaling in the newborn gut. Protection from NEC requires activation of AHR in the intestinal epithelium which is reduced in mouse and human NEC, and is independent of leukocyte activation. Finally, we identify an AHR ligand ("A18") that limits TLR4 signaling in mouse and human intestine, and prevents NEC in mice when administered during pregnancy. In summary, AHR signaling is critical in NEC development, and maternally-delivered, AHR-based therapies may alleviate NEC.

Fig. 1. Administration of the AHR ligand I3C during pregnancy protects against NEC in the newborn offspring.

a Schematic illustrating the experimental setup in which administration of the AHR ligand I3C to the pregnant mother can be evaluated for effects on the development of NEC in the pups. b–d mRNA expression of AHR activation marker Cyp1a1 in the ileum of pregnant mice (b, n = 3, 3 mice, p = 0.0003), the fetus (c, n = 9, 8 mice, p = 0.0317), and newborn mice at p1 (d, n = 3, 3 mice, p = 0.0053) or neomates at p11 (e, n = 5, 7 mice, p = 0.0002), on control diet (Ctrl) or maternal I3C enriched diet (M-I3C, 25 mg per kg body weight per day). f–i representative H&E-stained micrographs (f); NEC severity (g, n = 7, 6 mice, p = 0.0031); mRNA expression of Il6 (h, n = 13, 13 mice, p = 0.0329) and Tnf-α (i, n = 13, 13 mice, p = 0.0274) in the terminal ileum of pups with NEC from mothers (i.e., dams) who were fed either a control diet (Ctrl) or a diet rich in I3C (M-I3C, 25 mg per kg body weight per day). j–l AHR activation measured as Cyp1a1 expression in the intestinal epithelial cell line (IEC-6) treated with amniotic fluid (j, n = 11, 11 wells of cells, p = 0.0104), breast milk (k, n = 3, 3 wells of cells, p = 0.0002), and serum (l, n = 13, 12 wells of cells, p = 0.0007) that were harvested from the pregnant mice that were either on the control diet (Ctrl) or I3C enriched diet (M-I3C, 25 mg per kg body weight per day); m, n qRT-PCR showing expression of Il6 (m, n = 7, 8 mice, p = 0.0192) and Tnf-α (n, n = 7, 8 mice, p = 0.0004) in the ileum induced by LPS (50 μg per mL for 6 h) injection in neonatal mice, born from mice fed on control diet (Ctrl) or maternal I3C enriched diet (M-I3C, 25 mg per kg body weight per day) during pregnancy. Scale bars in f, 100 μm. All data are presented as mean values ±  SEM. *p < 0.05, **p < 0.01, ***p < 0.001, p values obtained from two-sided t tests. Each dot in graphs represents data from an individual mouse, or an individual well of cell culture.

Fig. 2. AHR expression on the the intestinal epithelium is required for protection against NEC.

a, b, d, e, g, h Representative confocal images and quantification of fluorescent intensity of AHR immuno-stained ileal sections of human (a, b, n = 7, 10 ileal section, p < 0.0001), mouse (d, e, n = 5, 5 ileal section, p = 0.0025), piglet (g, h, n = 7, 6 ileal section, p = 0.0004) ileal specimens from control (Ctrl) and NEC patients or animals. AHR, green signal; nuclei (DAPI, blue signal). c, f, i Expression of Ahr by qRT-PCR in the small intestine of human (c, n = 15, 9 human specimens, p = 0.0019), mouse (f, n = 8, 9 mice, p < 0.0001), and piglets (i, n = 5, 5 piglets, p = 0.0010). j H&E-stained representative images showing abnormal NEC histology in mice induced to develop experimental NEC in wild-type (WT), AHR knockout (Ahr^-/-), AHR intestinal epithelial cells knockout (Ahr^ΔIEC), and AHR myeloid knockout (Ahr^Δlys) mice. k Quantification of NEC severity (n = 7, 7, 7, 7, 7 mice, WT Ctrl vs WT NEC, p < 0.0001, WT NEC vs Ahr^-/- NEC p = 0.0074, WT NEC vs Ahr^ΔIEC NEC p = 0.0074, Ahr^-/- NEC vs Ahr^Δlys NEC p = 0.0001, Ahr^ΔIEC NEC vs Ahr^Δlys NEC p = 0.0001)^. l, m qRT-PCR expression of Il6 (l, n = 6, 14, 7, 9, 8 mice, WT Ctrl vs WT NEC, p = 0.0048, WT NEC vs Ahr^-/- NEC p = 0.0006, WT NEC vs Ahr^ΔIEC NEC p = 0.0268, Ahr^-/- NEC vs Ahr^Δlys NEC p = 0.0009, Ahr^ΔIEC NEC vs Ahr^Δlys NEC p = 0.0415) and Tnf-α (m, n = 6, 14, 7, 9, 8 mice, WT Ctrl vs WT NEC, p = 0.0350, WT NEC vs Ahr^-/- NEC p = 0.0012, WT NEC vs Ahr^ΔIEC NEC p = 0.0076, Ahr^-/- NEC vs Ahr^Δlys NEC p = 0.0473) in the intestinal epithelium. Scale bars in a, d, g, 50 μm. Scale bars in j, 100 μm. All data are presented as mean values ±  SEM. *p < 0.05, **p < 0.01, ***p < 0.001, p values obtained either from two-sided t tests or one-way ANOVA followed by multiple comparisons. Each dot in graphs represents data from an individual ileal section, human specimen, mouse, or piglet.

Fig. 3. Feeding mice the AHR ligand I3C protects against NEC by activating AHR on the intestinal epithelium.

a H&E-stained representative images in ileal sections of newborn mice, supplemented with AHR ligand I3C (25 mg per kg body weight per day for 4 days) and induced to develop experimental NEC. Data showing I3C mediated protection against NEC development only in wild-type (WT) and AHR myeloid knockout (Ahr^Δlys) mice but not in AHR knockout (Ahr^-/-) and AHR intestinal epithelial cells knockout (Ahr^ΔIEC) mice. b Dot graph showing AHR ligand I3C supplementation (25 mg per kg body weight per day for 4 days) produced a multifold induction of AHR activation marker Cyp1a1 in the ileum of wild-type but not in Ahr^-/-, and Ahr^ΔIEC and a moderate increase in Ahr^Δlys mice (n = 5, 11, 13, 11, 11 mice, Ctrl -I3C vs WT + I3C p < 0.0001, WT + I3C vs Ahr^Δlys + I3C p < 0.0001, Ahr^-/- + I3C vs Ahr^ΔIEC + I3C p = 0.0002, Ahr^ΔIEC + I3C vs Ahr^{Δlys +} I3C p = 0.0104). c–e NEC severity (c, n = 7, 7, 5, 7, 7 mice, Ctrl +I3C vs WT NEC + I3C p = 0.0020, WT NEC + I3C vs Ahr^-/- NEC + I3C p < 0.0001, WT NEC + I3C vs Ahr^ΔIEC NEC + I3C p < 0.0001, Ahr^ΔIEC NEC + I3C vs Ahr^Δlys NEC + I3C p = 0.0402) and mRNA levels of pro-inflammatory cytokine Il6 (d, n = 10, 16, 5, 10, 6 mice, WT NEC + I3C vs Ahr^-/- NEC + I3C < 0.0001, WT NEC + I3C vs Ahr^ΔIEC NEC + I3C p = 0.0001, Ahr^ΔIEC NEC + I3C vs Ahr^Δlys NEC + I3C p = 0.0086) and Tnf-α (e, n = 10, 16, 5, 10, 6 mice, WT NEC + I3C vs Ahr^-/- NEC + I3C p < 0.0001, WT NEC + I3C vs Ahr^ΔIEC NEC + I3C p < 0.0001, Ahr^ΔIEC NEC + I3C vs Ahr^Δlys NEC + I3C p = 0.0004) in the ileum of control mice without NEC and wild-type, Ahr^-/-, Ahr^ΔIEC, and Ahr^Δlys mice with NEC with I3C supplementation (25 mg per kg body weight per day for 4 days). Scale bars in a, 100 μm. All data are presented as mean values ± SEM. *p < 0.05, **p < 0.01, ***p < 0.001, p values obtained from two-sided t tests or one-way ANOVA followed by multiple comparisons. Each dot in graphs represents data from an individual mouse.

Fig. 4. AHR activation limits TLR4 signaling and expression in the intestinal epithelium of mice and humans.

a Representative confocal images of enteroids harvested from wild-type (WT) and AHR knockout (Ahr^-/-) mice after 7 days of culture on Matrigel, and stained with the epithelial marker (Ecadherin, Ecad, green signal), actin filaments (Rhodamine Phalloidin, RP, red signal), and nuclei (DAPI, blue signal). b, c qRT-PCR showing the expression of Cyp1a1 (b, n = 3, 3, 3, 3 wells of enteroids, WT -I3C vs WT + I3C p = 0.0208) and Tnf-α (c, n = 3, 3, 3, 3 wells of enteroids, Ctrl vs WT LPS p < 0.0001, WT LPS vs WT LPS + I3C p = 0.0258) in these enteroids under the indicated condition (200 μM I3C pretreatment overnight and then LPS treatment (50 μg per mL) for 4 h). d, e Representative confocal images (d) and quantification of fluorescent intensity of NF-κB translocation (e, n = 81, 48, 48, 50 enteroid cells, Ctrl -I3C vs LPS -I3C p < 0.0001, LPS -I3C vs LPS + I3C p < 0.0001). NF-κB, green signal; actin filaments (Rhodamine Phalloidin, RP, red signal); nuclei (DAPI, blue signal). f–i qRT-PCR showing the expression of Cyp1a1 (f, n = 6, 8, 9, 8 mice, WT -I3C vs WT + I3C p = 0.0042), Il6 (g, n = 6, 7, 8, 8, 8 mice, Ctrl vs WT LPS p = 0.0018, WT LPS vs WT LPS + I3C p = 0.0065), Tnf-α (h, n = 6, 7, 8, 8, 8 mice, Ctrl vs WT LPS p = 0.0036, WT LPS vs WT LPS + I3C p = 0.0017) and Tlr4 (i, n = 18, 19, 9, 8 mice, WT -I3C vs Ahr^-/- -I3C p = 0.0013, WT -I3C vs WT + I3C p = 0.0033) in the ileum of newborn mice after treatment with or without I3C (25 mg per kg body weight per day for 4 days). j, k qRT-PCR showing the expression of the Tlr4 regulatory microRNAs, miR-146b (j, n = 7, 5, 8, 5 mice WT -I3C vs WT + I3C p = 0.0300), let-7i (k, n = 6, 5, 7, 5 mice WT -I3C vs WT + I3C p = 0.0016) and miR-223 (l, n = 7, 5, 8, 5 mice WT -I3C vs WT + I3C p < 0.0001) in the ileum of WT and Ahr^-/- mice in the presence or absence of I3C (25 mg per kg body weight for 24 h). m, n qRT-PCR showing expression of CYP1A1 (m, n = 3, 3, 3, 3 wells of human explant culture, Ctrl -I3C vs Ctrl +I3C p = 0.0147, LPS -I3C vs LPS + I3C p = 0.0096) and TNF-α (n, n = 3, 3, 3, 3 wells of human explant culture, Ctrl -I3C vs LPS -I3C p = 0.0003, LPS -I3C vs LPS + I3C p = 0.0218) in the presence of I3C (200 μM I3C pretreatment for 15 min and then additional 6 h) and LPS (50 μg per mL for 6 h). Scale bars in a, 25 μm. Scale bars in d, 10 μm. All data are presented as mean values ±  SEM. *p < 0.05, **p < 0.01, ***p < 0.001, p values obtained either from two-sided t tests or one-way ANOVA followed by multiple comparisons. Each dot in graphs represents data from an individual well of enteroids culture, an individual enteorid cell of NF-κB staining, an individual mouse, or an individual well of human explant culture.

Fig. 5. Breast milk activates AHR on the intestinal epithelium and protects against experimental NEC in newborn mice.

a, b qRT-PCR showing the expression of Cyp1a1 (a, n = 6, 5, 6, 5 wells of enteroids, WT saline vs WT breast milk p = 0.0096) and Tnf-α (b, n = 5, 3, 3, 3, 3 wells of enteroids, Ctrl vs WT LPS p = 0.0001, WT LPS vs WT LPS + breast milk p = 0.0325) in mouse intestinal enteroids from wild-type (WT) and Ahr^-/- mice, treated with saline or LPS in the presence or absence of breast milk. c, d LPS-induced apoptosis in enteroids measured by TUNEL assay in wild-type and Ahr^-/- mouse enteroids as quantified (c, n = 7, 5, 7, 10, 4 enteroid sections, Ctrl vs WT LPS p < 0.0001, WT LPS vs WT LPS + breast milk p < 0.0001) and revealed (d) by confocal microscopy. TUNEL, green signal; nuclei (DAPI, blue signal). e–g qRT-PCR showing the expression of Cyp1a1 (e, n = 9, 6, 9 5 mice, WT saline vs WT breast milk p = 0.0026), Il6 (f, n = 9, 6, 4, 5, 3 mice, Ctrl vs WT LPS p = 0.0039, WT LPS vs WT LPS + breast milk p = 0.0070) and Tnf-α (g, n = 5, 3, 3, 3, 3 mice, Ctrl vs WT LPS p = 0.0001, WT LPS vs WT LPS + breast milk p = 0.0325) in wild-type and Ahr^-/- mice treated with saline, LPS and/or breast milk as indicated. h–l Representative H&E- stained section (h), NEC severity score (i, n = 6, 7 mice, p = 0.0023), and the expression of Il6 (j, n = 7, 6 mice, p = 0.0056), Tnf-α (k, n = 7, 6 mice, p = 0.0221) and Tlr4 (l, n = 7, 6 mice, p = 0.0465) in wild-type and Ahr^-/- mice induced to develop NEC in the presence or absence of breast milk as indicated. Scale bars in d, 10 μm. Scale bars in h, 100 μm. All data are presented as mean values ±  SEM. *p < 0.05, **p < 0.01, ***p < 0.001, p values obtained either from two-sided t tests or one-way ANOVA followed by multiple comparisons. Each dot in graphs represents data from an individual well of enteroids, section of enteroids, or an individual mouse.

Fig. 6. The AHR agonist “A18” activates AHR on the intestinal epithelium, reduces TLR4 signaling, and protects against experimental NEC in newborn mice.

a Molecular formula of A18. b Dose-response curve of A18 for luciferase activity in AHR-reporter IEC-6 cells (n = 6, 2, 2, 2, 2, 2, 2, 2 wells of cells, 0 vs 2.5 p = 0.0390, 0 vs 5 p = 0.0081, 0 vs 10 p = 0.0006,). c qRT-PCR showing expression of Cyp1a1 in IEC-6 cells treated with A18 (20 μM for 6 h) (n = 4, 4, wells of cells, p = 0.0018). d–g qRT-PCR showing expression of Cyp1a1 (d, n = 8, 8, 8, 8 mice, WT - A18 vs WT + A18 p < 0.0001), Il6 (e, n = 9, 4, 5 mice, Ctrl vs LPS p < 0.0001, LPS vs LPS + A18 p < 0.0001), Tnf-α (f, n = 9, 4, 5 mice, Ctrl vs LPS p = 0.0003, LPS vs LPS + A18 p = 0.0017) and Tlr4 (g, n = 9, 4, 5 mice, Ctrl vs LPS p = 0.0004, LPS vs LPS + A18 p = 0.0011) in the ileum of wild-type (WT) but not AHR knockout (Ahr^-/-) mice. h–k Representative histological H&E of the terminal ileum (h), NEC severity score (i, n = 6, 7 mice, p = 0.0003), expression of Il6 (j, n = 5, 7 mice, p = 0.0484) and Tnf-α (k, n = 5, 7 mice, p = 0.0222) in newborn mice with NEC after administration of A18 (300 mg per kg body weight per day for 4 days). l–n qRT-PCR showing expression of CYP1A1 (l, n = 3, 3 wells of human explant culture, p = 0.0106), TNF-α (m, n = 3, 3 wells of human explant culture, p = 0.0035) and TLR4 (n, n = 3, 3 wells of human explant culture, p = 0.0001) in freshly harvested human ileum obtained at surgery treated with LPS (50 μg/mL for 6 h) and/or A18 (20 μM A18 pretreatment for 15 min and then additional 6 h). o–t qRT-PCR showing expression Cyp1a1 (o, n = 7, 6 mice, p = 0.0377) in the neonatal ileum after maternal administration of A18 (300 mg per kg body weight per day for 4 days), expression of Il6 (p, n = 7, 6 mice, p = 0.0254), Tnf-α (q, n = 7, 6 mice, p = 0.0346) and Tlr4 (r, n = 7, 6 mice, p = 0.0081), ileal H&E of the newborn (s) and NEC severity (t, n = 7, 6 mice, p = 0.0023) after NEC induction in the absence or presence of maternal A18 (M-A18, 300 mg per kg body weight per day for 4 days). Scale bars in h, s, 100 μm. All data are presented as mean values ±  SEM. *p < 0.05, **p < 0.01, ***p < 0.001, p values obtained either from two-sided t tests or using one-way ANOVA followed by multiple comparisons. Each dot in graphs represents data from an individual of cell culture, an individual mouse, or an individual well of human explant culture.